Transcript Studies of a Proximity focusing RICH with Aerogel Radiator

International Workshop on New PhotoDetectors (PD07) @ Kobe University
Geiger-mode APD as
a RICH Photodetector
Toru Iijima
Nagoya University
June 28, 2007
Collaborators
Koji Hara, Toru Iijima, Yuri Mazuka, Mio Yamaoka
Department of Physics, Nagoya University
Rok Dolenec, Samo Korpar, Peter Krizan, Rok Petotnik,
Andrej Petelin
Jozef Stefan Institute, Ljubljana
Special Thanks to;
Belle-ACC group, KEKDTP RD-photon group
Hamamatsu Photonics, Yuri Kudenko
2007/6/28
Toru Iijima, PD07 @ Kobe
2
Talk Outline
Can we use Geiger-mode APD (SiPM / MPPC)
as a RICH photosensor ?






RICH w/ Aerogel radiators
Why Geiger-mode APD for RICH ?
Characterization of G-APD
Light collection system
Expected performance
Summary/Prospect
2007/6/28
Toru Iijima, PD07 @ Kobe
3
Proximity Focusing Aerogel RICH

Aerogel radiator (n~1.05, ~2cm)
+ photodetector (Dx ~ 5mm)

>4s K/p for 0.7 < p < 4.5 GeV/c

Proximity focusing geometry
No mirror complex.
 Suitable for collider and space
experiments.


Rayleigh scattering dominates in
aerogel.
 Demand for positioning of a
single photon in the visible wave
length region.
2007/6/28
Design values for Belle upgrade
Toru Iijima, PD07 @ Kobe
Npe
~7.5
s0
11mr
s（pix）
6.4mr
s（em）
8.6mr
s（chr）
2.0mr
4
Belle PID Upgrade Option


Barrel  TOP (Time-Of-Propagation) Counter
Endcap  Proximity Focusing Aerogel-RICH
Barrel PID  TOP
Endcap-PID
 Aerogel-RICH
Talk by K.Inami
Photodetector options
 HAPD (baseline)
Poster by I.Adachi

MCP-PMT
Talk by S.Korpar

2007/6/28
Toru Iijima, PD07 @ Kobe
G-APD
5
RICH with Multiple Radiators
NIM A548(2005)383

Demonstration of principle

4×4 array of H8500 (85% effective area)
Conventional
sc=22.1mrad
Npe=10.7
4cm thick aerogel
n=1.047
Multiple
Radiator
s layers of 2cm thick
2
sc=14.4mrad
Npe=9.6
n1=1.047, n2=1.057
p/K separation with focusing configuration ~ 4.8s @4GeV/c
6
2007/6/28
Toru Iijima, PD07 @ Kobe
RICH with Precision Timing
Fast photon detector enables Aerogal RICH to have TOF info.
In case of Belle…
photosensor
DTring(p/K) ~37ps
aerogel
p/K 4GeV
DTwindow(p/K) ~47ps
IP
1.8m
0.2m
Beam test result
w/ Burle 85011-501
p/K separation
w/ TOF
swindow = 34 ps
w/ glass hit
Positive ID of
K/p below
thereshold.
2007/6/28
@ 4GeV/c
Toru Iijima, PD07 @ Kobe
7
Why SiPM/MPPC for RICH ?

Comparison to other photodetctors
PMT
MCP-PMT
HPD / HAPD
G-APD
Gain
>106
～106
～103
X10～100 w/ APD
～106
Quantum Eff.
～20%, ～400nm (bialkali)
～80%, ～600nm
to be checked
Collection Eff.
70%
60%
100%
50%
Time
resolution
～300ps
～30ps
～150ps
<100ps
Depends on readout
To be checked
B-field
immunity
×
Problems
△ Depends on angle
○
lifetime
Noise, size
Pros.
Cons.



High PDE
B field immunity


2007/6/28
Noize
Size
Rad. hardness ?
Toru Iijima, PD07 @ Kobe
8
Possibility of G-APD for RICH
To improve S/N …
 Increase the number of signal hits/sensor by using light
collectors (with optimization of pad size to the ring thickness).
 Reduce the noise by a narrow time window (<10ns).
Light collector serves as a light emitter for TOF measurement.
SiPM / MPPC + Optics (Light-guide. Lens…)
Aerogel-RICH
Cherenkov Photons
(17deg max. for n=1.05)
n1 n2
G-APD
□1~3mm
Light collector
□3~5mm (IN)
□1-2mm(OUT)
G-APD Characterization
New
Measured Items


Old
Photon-counting performance
Timing property
20mm
Use of timing for random noise rejection
 + also for TOF measurement


Photo-Detection Efficiency


As a function of l
Noise Performance
We report basic performance using
1 x 1 mm2 samples
Measured samples
MPPC from HPK
Sample
H100old
H100new1,2
H400new1,2
1.0x1.0mm2
Area
Production.
H400old
2005 / beg.
2006 /end.
#pixels
100
400
100
400
Geom. Eff. (%)
-
-
78.5
61.5
+
MRS-APD (CPTA)
1710 series
1.1mm2
556 pixels.
SiPM (MePhI/PULSAR)
1710 series
1.1mm2
556 pixels.
Performance: Photon-counting

Excellent resolution to separate 0 and 1 photon (and more).

Gain = 1.8 x 106 at Vbias=71.5V

S/N = Dmean/s = 12
H100-old, Vbias=71.5V, noise～1.0MHz
meanped: 98.9
mean1pe: 137.1
sped: 3.0
H100-old.
Pulse laser
HPK PLP-02 (410nm)
ALS PiL063 (636nm)
Light dielding box
xy-stage
Performance: PDE
A
Photo Detection Efficiency
A

QE
 geom
 Geiger



Quantum efficiency
Geometrical efficiency
Geiger efficiency
Higher efficiency at peak and
long wave length, compared to
conventional PMT’s.
Our result is consistent with
HPK.
Need confirmation by photocounting to disentangle the
effects of cross talks and
after-pulses.
P.D.E. (%)
PDE  QE   geom   Geiger
Ref.
PD
Geiger
mode
APD
filter
pin hole
f200mm
45% at peak (460nm)
Photo
spectrometer
Performance: Gain vs Noise
Recent HPK products have much lower noise rate
< 1/3 at the same gain.
100 pixel
Noise (kHz)

2007/1/16
Gain (x 103)
400 pixel
Performance: Time Resolution
Measured w/ pulse laser
636 / 410nm
MPPC
(HPK)
Time walk
corrected.
MRS-APD
636nm
s~103ps
s~70ps
Sample ＭＰＰＣ
Bias
-71.5V
Threshold 0.5pe
Only Single photon data
410nm
s~110ps
s~140ps
Depend on internal structure ?
Can be used for TOF measurement as well.
Time resolution (MRS-APD)
Single photon, threshold ~0.4pe
l=635nm
l= 405nm
long tail(~4ns)
Short tail(~0.5ns)
s~69ps
39.5
2007/6/28
s~142ps
40.0 40.5 41.0 41.5
TDC(ns)
41.5
Toru Iijima, PD07 @ Kobe
42.0 42.5 43.0 43.5
TDC(ns)
15
Setup for Surface Scan
@ Jozef Stefan Inst.
2007/6/28
Toru Iijima, PD07 @ Kobe
16
SiPM surface sensitivity
Size: ~1mm
Scanned with laser, resolution ~5 mm
Single photon response
SiPM (MePHY/Pulsar)
MRS APD by CPTA (Moscow)
2007/6/28
Toru
Iijima,
PD07 @
Kobe
17
Micro Structure
SiPM (MePHY/Pulsar)
MRS APD by CPTA (Moscow)
2007/6/28
Toru
Iijima,
PD07 @
Kobe
18
Hamamatsu MPPCs
100 pixels
400 pixels
1mm
19
Light Guide

Possible design




Trapezoid
Winston Cone
Lens
…
Ex.) Trapezoid, incident at 0.3rad (max. in case of Super-B)
□5mm  □1mm
□5mm  □2mm
CE=95%
CE=37%
2007/6/28
Toru Iijima, PD07 @ Kobe
20
Simulation conditions
•
•
•
•
Light Guide


Trapezoid
Incident angle = 0.3rad
Wave length = 400nm
No internal absorption
n=1.47 acryl ic material
Lens (half-sphere)
+Trapezoid
length
5mm
length
L>12mm to obtain max. eff.
2mm
5mm
2mm
L>9mm to obtain max. eff.
Light Guide (2)
2007/6/28
Toru Iijima, PD07 @ Kobe
22
Arbitrary
Expected RICH Performance
The simulation is adjusted to
reproduce beam test results
with multi-anode PMT.
5mm
2mm
A photosensor based on Light-guide + Geiger-mode APD can give
x 4 Npe (number of detected photons).
Should be corrected for cross talk / after-pulse contribution
(because PDE based DC current measurement is used here).
Effect of Noise


Random background embeded in a simulation.
Assume Npe = 20 / ring (conservatively)
Kaon ID efficiency at 1% pion mis-ID probability
0.8%
1.6%
3.2%
0.8%
1.6%
3.2%
Pad size = 4mm2
Momentum (GeV/c)
Momentum = 4 GeV/c
Pad size (mm2)
Noise rate (HPK-100) = 200KHz/mm2 @ 106 gain.
0.8MHz for 4mm2 pad and 10ns time window
Summary




Geiger-mode APD is very attractive device as a photosensor for a
RICH with aerogel radiator.
 High PDE  Significant increase in Npe (x 4 possible).
 Good time resolution (s~100ps /p.e.)  RICH w/ TOF
 Free from magnetic field  Large advantage !
Noise rate has been reduced significantly and in a tolerable region
(<1MHz/pad) for recent products from HPK .
 Very encouraging !
Light colletction based on tapered lightguide (+lens) will be useful to
increase the detection area, and hence improve S/N.
Devices with larger size (□3~5mm) are highly welcome.
We are starting to measure □3mm sample from HPK.
2007/6/28
Toru Iijima, PD07 @ Kobe
25
Backup
2007/6/28
Toru Iijima, PD07 @ Kobe
26
Light absorption in silicon
Electric field in SiPM
Light absorption length a = f (l)
l (nm)
a(mm)
s(ps)
tail(ns)
410
0.16
140
~0.5
639
3.65
70
~4
This is plot of SiPM, not MRS-APD
reference: ICFA Instrum.Bull.23:28-41,2001
Red(l=635nm) :
absorbed in Geiger region
→make prompt signals
absorbed in drift region
→make slow signals, long tail
g
Blue(l=405nm) :
absorbed before reach to Geiger region
→TDC has short tail,
g
worse resolution in Geiger region
27/9
2007/6/28
(mm)
e h
e h
- e h
-
n+
p+
e h
-
pe h
-
e h
Toru
Iijima,
PD07 @
Kobe
Temperature Dep. of Noise
MRS-APD
MPPC
Geiger mode APDが並んでいる
青リング チェレンコフリング
赤 チェレンコフ光による信号
緑 ノイズによる信号
（deg.）

０℃では室温の20～30％になった
ＭＲＳ－ＡＰＤ
ＭＰＰＣ
～300ｋＨｚ@ 0℃
～200ｋＨｚ @ 0℃
Noise (kHz)
Performance: Gain vs Noise (MRS-APD)
Gain (x 103)
Plan




Further measurements of samples
 HPK starts to provide 3x3mm2 samples.
 Better understanding of critical properties.
 Feedback to producers.
Light-guide design
 Measurements with a test sample.
Photo from the talk by Sato-san (HPK)
at the photosensor WS (Dec.7-8,2006).
Method to produce the assembly.
 How to make the light-guide (LG) array.
 Optical connection between LG and SiPM/MPPC.
Readout Electronics